The Idea Factory: Bell Labs and the Great Age of American Innovation
Page 7
In Kelly’s research department on West Street, Shockley found he could go mostly where his curiosity led him, which was often to solid-state physics. He likewise found that at the Labs the experimentalists and theoreticians were encouraged to work together, and that chemists and metallurgists were welcome to join in, too. The interactions could be casual, but the work was a serious matter. Every new member of the technical staff was given a stock of hardcover lab notebooks that were bound in cloth and leather and filled with two hundred lined pages. In most offices, recalls Walter Brown, an experimental physicist who worked under Shockley, there was a notebook table, “maybe twelve by eighteen inches, standing on a three-legged stand on the floor, painted black. It was intended to hold a notebook for recording details of experiments and their results [as well as] ideas and plans for the future. Results or ideas that one thought were potentially valuable were witnessed and signed by another engineer for documentation of the timing of the idea.” The scientists were not permitted to rip out pages. Nor were they encouraged to attach loose sheets of paper into the notebook. “No erasures,” says Brown. “Lines through mistakes—initialed by who drew the lines.”18 Also, the notebooks were issued with registered numbers that were matched to each scientist and were tracked by supervisors and Labs attorneys. There was to be no confusion about who did what. The notebooks were proof for gaining a patent.
At some point in late 1939, Shockley had settled on an idea for how to make an electronic amplifier—much like the old repeater tube that Harold Arnold had improved—but this time out of solid materials. The production of vacuum tubes had improved since the days when Kelly ran the tube shop, but the essential problems remained: They were still fragile, they still consumed much electricity and produced much heat. The first attempts at making a solid-state amplifier, as Shockley was trying to do, involved simply copying the architecture of vacuum tubes. Shockley recalled later that his “first notebook entry on what might have been a working [solid-state amplifier] was as I recall late 1939.”19 It was actually December 29, 1939. Shockley had concluded by then that a certain class of materials known as semiconductors—so named because they are neither good conductors of electricity (like copper) nor good insulators of electricity (like glass), but somewhere in between—might be an ideal solid replacement for tubes. Under certain circumstances semiconductors are also known to be good “rectifiers”; that is, they allow an electric current passing through them to move in only one direction. This property made them potentially useful in certain kinds of electronic circuits. Shockley believed there could be a way to get them to amplify a current as well. He intuited that one common semiconductor—copper oxide—was a good place to start.
As a physicist, Shockley was far better as a theoretician than an experimentalist. On the other hand, Walter Brattain, his colleague at West Street, was about as good an experimentalist as could be found at Bell Labs. With good reason, Brattain prided himself on being able to build anything. “He came to me one day and said that he thought that if we made a copper-oxide rectifier in just the right way, that maybe we could make an amplifier,” Brattain recalled. “And I listened to him. I had a good esprit de corps with him, and so after he explained, I laughed at him.” Brattain, it turned out, had already tried a variation on the idea with another colleague. But when he saw how intent Shockley was on trying out his idea, Brattain went along, pledging that he would make a prototype to Shockley’s precise specifications. In the early winter months of 1940, Brattain built a couple of units to Shockley’s specifications. “It was tested and the result was nil,” he recalled. “I mean, there was no evidence of anything.”20
But Shockley wasn’t convinced his idea was wrong. He would speculate later about what might have occurred had he continued to develop that particular amplifier experiment without interruption. But as it happened he couldn’t. In fact, few people at the Labs could carry on their customary work anymore. The news from Europe—beginning with Germany’s invasion of Poland in 1939, and its invasion of Belgium, France, and the Netherlands in the spring of 1940—put an end to business as usual.
Four
WAR
By the middle of 1940, the research department at Bell Labs stopped doing research as nearly all of the Labs’ work—about 75 percent of it—was redirected toward developing electronic devices for wartime, first to help the Allies in Europe, and soon after to assist the U.S. Army and Navy. Frank Jewett, Bell Labs’ president, began spending almost all his time in Washington advising political leaders on how the country’s scientists might contribute to the war effort. Oliver Buckley, the Labs’ vice president, mostly focused on their obligation to maintain phone company operations. “Buckley in essence handed over Bell Laboratories to [Mervin] Kelly during World War II,” one Bell Labs researcher recalled.1 Indeed, Buckley let Kelly—whose research department was essentially dismantled for the duration of the war and replaced by a multitude of development projects—run the day-to-day business of the laboratories from that point on.
One of the Labs’ first assignments as the war began in Europe resulted from Jewett’s political connections:2 finding out at the government’s behest whether it was actually possible, in light of several recent theoretical papers on nuclear reactions, to create a weapon out of ordinary uranium.3 At Kelly’s request, Shockley and Jim Fisk, his friend from MIT who had just joined the physics department at the Labs, were asked to take time off to prepare a report. Shockley did most of the calculations. And while the two men quickly concluded that uranium in its natural state could not make a devastating weapon, they theorized that by placing “piles” of a specially enriched uranium preparation close together one might be able to create a sustained, low-level reaction. Put simply, they’d figured out how to make a nuclear reactor.4 The men tried to patent the idea, but met with resistance from the government and the patent courts. As Fisk would recall, the reason was that the physicists Enrico Fermi and Leo Szilard “had essentially the same idea and probably at about the same time. We may have been earlier, they may have been earlier, I don’t know. I don’t think that anybody will ever know. But they were working hard on this and we were doing this simply as an exercise to answer the question.”5
Like Kelly, Shockley rarely lingered over any one project. That he had figured out the essential concepts for nuclear power on his own (actually, the idea came to him while he was taking a shower) merely seemed an intriguing interlude in a frenetic schedule. Indeed, his schedule was too busy for him to do anything else but keep moving. Beginning in 1940, Kelly assigned Shockley to a secret effort to help develop applications for a new technology known as radar. Other members of the technical staff were asked to put aside their research work, too, in the same way that Kelly and Davisson had done two decades earlier during World War I.
There was no use challenging the Labs executives on the matter. For one thing, it was Kelly giving the orders. He still carried forth the brusque manner that had defined him as a young man, but he was now a boss who wielded power forcefully, and sometimes fearsomely. For another thing, researchers in Kelly’s department rarely had any objections to pitching in on the war effort. One day not long after he began his war work, Charles Townes was walking through Times Square when a complete stranger came up to him: “You’re not in uniform. Shame! A man your age ought to be in uniform, and helping out.”6 Townes was working fifty- and sixty-hour weeks to do precisely that. To the general public, however—largely unaware that this war would depend as much on technology as strategy—their role remained obscure. For a while, at least, so did the devices they were working on.
ON THE HOME FRONT, the war overturned the established norms of science, engineering, and business. Mervin Kelly had long regarded scientific research as a pursuit of the unknown that inherently defied corporate and political regimes. Science had no true owners, only participants and contributors. By the early 1940s, the great discoveries of his lifetime, including the work of his friends Robert Millikan at Chicago and Davy D
avisson at Bell Labs, were such that they transcended borders and nationalities. The results of their work were widely shared, discussed, and augmented, as Kelly thought they should be, through international gatherings and cooperative, investigative efforts. Engineering, however, was different. Kelly defined it as the application of science to a problem affecting society. Engineers dipped into the “common reservoir” of science on behalf of their own industries and countries. In peacetime, that meant they focused on making profitable commodities like automobiles and telephone equipment; in wartime, that meant they focused on building military communications equipment as well as ships, planes, and munitions. At the same time, wartime engineers had an additional responsibility. They were charged not only with building everything better but building everything faster, which meant striving to improve their processes as well as their products. In the summer of 1943, Kelly wrote an article for an engineering publication called The Bridge that directed attention to the contributions of the American engineers working on the Allied war effort. In particular he remarked on the speed with which U.S. industry had caught up with the military economies of the Axis powers. “Progress has been made in some fields of technology in a four-year interval,” he pointed out, “that, under the normal conditions of peace, would have required from ten to twenty years.” Much of what Bell Labs was now building hadn’t existed four years before. “In this astonishing short period of time,” he added, “we have developed, designed, placed in manufacture and expanded to unprecedented high rates the production of a substantial portion of the tools of warfare that our Army and Navy now are so successfully employing.” If the peacetime needs of the rapidly expanding phone system were an unceasing force for new inventions, in other words, war was turning out to be an even greater force.
One obvious reason was that war created an air of urgency. An example Kelly liked to cite was the supply of quartz crystals for telephone filters—the crucial equipment used to scramble and unscramble transmissions at either end of a cable.7 In the late 1930s, the supply of quartz from South America slowed noticeably. The countries of the world had already begun competing for valuable resources. “We had been working in the period before the war in growing crystals in the laboratory,” Kelly explained, “growing out of solutions crystals which would have the same electrical and mechanical properties as quartz.” War accelerated these developments. Technicians in fact discovered that by placing a small “seed” in a vat of chemicals—the scientists tried a hundred different formulas before settling on one—artificial crystals could be grown to the length of six inches, resembling in their glass tanks huge clusters of rock candy. At that point the fully grown crystals were removed from the tanks, the New York Times explained, “sawed into thin wafers, ground carefully to precise thicknesses, mounted in special holders and installed in electrical circuits.”8 It was a good example, in Kelly’s view, of how America’s scientists and engineers had responded to the war effort. The Labs ultimately produced hundreds of thousands of synthetic crystals.
The war also unleashed a great gush of money toward new technology. In an effort to quicken the development of equipment as much as possible, beginning around 1941 the federal government began directing hundreds of millions of dollars toward engineering organizations like Western Electric and Bell Labs. In the first few years after Pearl Harbor, in fact, Bell Labs took on nearly a thousand different projects for the military—everything from tank radio sets to communications systems for pilots wearing oxygen masks to enciphering machines for scrambling secret messages—leading Kelly to expand his staff by several thousand. The Labs actually doubled its size from about forty-six hundred before the war to nine thousand during it. At the West Street offices, Oliver Buckley wrote, “there is hardly room to turn around.” Elevators were so jammed with employees that it was difficult to squeeze on. A six-day workweek became the norm.
Not all the new employees fit the profile of the old-time telephone engineer. Some seven hundred Bell Labs staff members had gone into active military service. (“There are so many Laboratories people on Guam at one time or another,” the Bell Laboratories Record noted, “that they might have opened a Western Pacific branch.”) This led Labs executives to hire hundreds of women to replace the men. What’s more, for the first time, the Labs began to hire Jews, bucking a strain of anti-Semitism that ran deep within the AT&T establishment, though not, apparently, within Kelly. The precise reasons for the shift remained ambiguous, though some Labs members later reflected that by 1940 the specter of war had trumped the Bell System’s ugly traditions of bigotry. A slightly different explanation was that a meritocratic organization such as the Labs could perceive a competitive disadvantage of passing over the best scientists on religious grounds, an error they might have already made with the young Richard Feynman, a former colleague of Shockley’s at MIT who would eventually be drafted into the Manhattan Project.9 Whatever the explanation, some of the older and most hidebound scientists at the Labs found discomfort in this aspect of the Labs’ evolution, as well as in Kelly’s war mobilization effort. Lloyd Espenschied, for example, an inventor of the thick coaxial cables that carried phone conversations between major cities and a senior advisor to Oliver Buckley, was an avowed isolationist. In the midst of the war, an investigator for the War Department named G. E. Schwartz visited Espenschied at West Street. “We were led into this mess largely by British propaganda, Jewish propaganda, Roosevelt imperialism,” Espenschied told him in response to his questions.
He then said “I suppose you are anti-Semitic.” I told him he could take that supposition if he wished, that frankly I do not like Jews. He then asked me why I did not like them and I told him that it was because of their racial characteristics. To his question of why was this, I responded that he would have to ask the Jews themselves that question, looking him straight in the face as the Jew that he appeared to be. I told him that to my mind all this was part of the problem as a result of different peoples being thrown together rather rapidly in this age of increased mobility and shrinking distances.10
Frank Jewett and Oliver Buckley were appalled by Espenschied’s “incredible stupidity”—though it was not clear whether their displeasure related to the content of Espenschied’s opinions or his willingness to share them. “His error lies in his insistence on expressing his views intemperately,” Jewett noted in a private memo about the exchange.11 Espenschied was forbidden from having any contact with work on the war, or with any Labs employees involved with such work. He was not dismissed, however—a report on the incident was filed away by Buckley and stamped as “Confidential F.B.I.” Mostly it seemed that Jewett and Buckley were primarily concerned that one of their most accomplished engineers might be a blemish on the Labs’ reputation for patriotism and national service.12
Probably they worried needlessly. Thousands of men and women under Kelly were working intently on military applications, and the public was increasingly aware of their contributions. Press coverage of various aspects of what was known as “the physicists’ war” was glowing. Still, for many members of the technical staff, the wartime work required a difficult philosophical transformation. The ideas of scientists thrive on publication and broad dissemination; but the ideas of engineers, especially during wartime, thrive only if secrecy is maintained. For the first few years of World War II, the word “radar”—its name stood for “radio detection and ranging”—was almost unknown by the general public, earning it the description of a New York Times reporter as being “the war’s most fabulously and zealously guarded secret.”13 This was not quite true—the Manhattan Project was kept under tighter wraps—but it was generally the case that Kelly’s men at Bell Labs, many of whom worked on radar systems, were forbidden to speak with anyone, even their wives, about their work, and that virtually all details about radar and its sister technology for underwater detection, sonar, were closely held for the duration of the war. Harvey Fletcher, for instance, who contributed his knowledge of acoustics to the Bell Labs
sonar work during the early 1940s, actually refrained from speaking of his involvement for fifty years after victory was declared. Kelly never offered any details about his war work and likely either destroyed or purposely discarded his personal files on military matters. Secrecy began to cloak his responsibilities, as well as his power. When a three-hundred-page internal history of Bell Labs’ World War II work was later compiled, his name was never mentioned.
THE SCIENTIFIC PRINCIPLES OF RADAR were fairly straightforward, even if the details of the technology weren’t. A memo to Bell Labs employees explained that radar could be defined as a powerful electronic “eye” that used high-frequency radio echoes to determine the presence and location of unseen objects in space: “Specifically, a radar system does the following: (1) it generates high-power electrical waves; (2) it projects these waves from an antenna, usually in a narrow beam; (3) it picks up the waves which reflect back from objects in its range; and (4) converts these into a pattern on a fluorescent screen.” The memo might have added that the waves traveled and bounced back at the speed of light, about 186,000 miles per second. Thus, in evaluating the time it took for radar waves to leave and then echo back to a radar antenna, a set could also calculate the distance of an unseen object based on the knowledge that distance equals rate multiplied by time. (Radar equipment was designed to divide in half the distance a wave traveled, since operators didn’t need the distance to and from the object, only the distance in one direction.) It all happened instantaneously. An object one thousand yards from the radar set would give its echo six-millionths of a second later.14